Estimating Intersection Annual Average Daily Bicycle Traffic from 8-hour Turning Movement Counts

Abstract

This thesis outlines a set of procedures for estimating annual average daily bicycle traffic (AADB) from one day, 8-hour turning movement counts (TMCs). Factoring methods for annualizing short duration counts have been long established for motor vehicle traffic, and more recent research has adapted many of these methods to pedestrians and bicyclists. However, much of this research has been concerned with estimating AADB on dedicated cycling facilities. Less has been done to transfer these methods to estimating cyclist activity at intersections, even though it would be valuable to measure cyclist exposure for network safety analysis and for broader planning purposes. TMCs represent a valuable potential source of data for this purpose, as it is common for North American jurisdictions to regularly collect them as part of ongoing traffic monitoring programs. Sets of video monitoring unit (VMU) data from Milton, Ontario, and Pima County, Arizona, were used to evaluate whether existing methods could be appropriately applied to 8-hour TMCs. Several updates to conventional estimation methods were proposed to account for the differences between TMCs and “conventional” cyclist counts. Additionally, methods are proposed for filtering VMU data; and for matching short-duration count locations to empirical factor groups using their land-use and physical characteristics. The resulting set of procedures could be implemented by transportation agencies using data which they may already be collecting to generate estimates of cyclist activity at any intersection in their jurisdiction, although further work is likely needed to improve estimation accuracy, especially at low-volume locations.